1

Pathogenic and molecular diversity in highly virulent populations of the parasitic 1

weed Orobanche cumana (sunflower broomrape) from Europe 2

3

L MOLINERO-RUIZ*, A B GARCÍA-CARNEROS*, M COLLADO-ROMERO*, S 4

RARANCIUC†, J DOMÍNGUEZ‡ & J M MELERO-VARA* 5

6

* Institute for Sustainable Agriculture, CSIC, Apdo. 4084, 14080, Córdoba, Spain, † 7

NARDI, 915200 Fundulea, Romania and ‡ IFAPA Centro Alameda del Obispo, CAPMA 8

(Junta de Andalucía), Apdo 3092, 14080 Córdoba, Spain 9

10

11

Short title: Diversity in highly virulent Orobanche cumana from Europe 12

13

14

Correspondence: L. Molinero-Ruiz, Institute for Sustainable Agriculture, CSIC, Apdo. 15

4084, 14080 Córdoba, Spain. Tel: (+34) 957 499272; Fax: (+34) 957 499252; E-mail: 16

leire.molinero@csic.es 17

18

19

Código de campo cambiado

2

Summary 1

The parasitic weed Orobanche cumana (sunflower broomrape) constrains sunflower 2

production in eastern and southern Europe and in the Middle East. Although genetic 3

resistance is the most effective control method, new parasite races evolve overcoming 4

sunflower resistance. In this work, highly virulent populations of O. cumana were analysed 5

for pathogenicity and genetic diversity. The virulence of 11 populations from Hungary, 6

Romania, Spain and Turkey was assessed and compared after infection of sunflower 7

inbred lines to differentiate races of the parasite under glasshouse conditions. Molecular 8

diversity among and within 27 parasite populations was studied by RAPD-PCR, UPGMA 9

and AMOVA analyses. Highly virulent race F was identified in Hungary, Spain and 10

Turkey. The most virulent race (G) was also found in Turkey. The molecular analysis 11

among highly virulent populations of O. cumana identified four molecular clusters, 12

respectively grouping populations from Central Spain, Hungary, South Spain and Turkey. 13

The genetic homogeneity within parasite populations was confirmed, since no molecular 14

divergences were found within them. This work constitutes the first geographical study of 15

O. cumana together with pathogenicity and molecular traits inherent to each geographical 16

group, and provides useful information for possible phylogenetic analyses of O. cumana. 17

In addition, molecular markers associated with geographic origin could be developed and 18

used as diagnostic tools to track new broomrape introductions into areas free of virulent 19

races where they might represent a threat to sunflower production. 20

21

22

Keywords: AMOVA analysis, genetic resistance, genetic diversity, Helianthus annuus, 23

molecular characterisation, pathogenicity, RAPD-PCR analysis. 24

25

3

Introduction 1

Sunflower (Helianthus annuus L.) is the most important annual oilseed crop in southern 2

Europe and the Black Sea area. The holoparasitic angiosperm Orobanche cumana Wallr. 3

(sunflower broomrape) is regarded as one of the most important constraints for crop 4

production in most countries of the Middle East and eastern and southern Europe, 5

including Hungary, Romania, Spain and Turkey. Orobanche cumana infects sunflower 6

roots and depletes the plant of water and nutrients causing up to 50% of yield losses in 7

susceptible cultivars (Dominguez, 1996). Soil fumigation and solarization are not 8

economically feasible (Foy et al., 1989; Jacobson et al., 1980) and imidazolinone 9

herbicides can be very effective, but they must be applied to imidazolinone-resistant 10

sunflowers (Clearfield production system) (Tan et al., 2005). Genetic resistance is still the 11

most frequent, feasible and reliable control method against sunflower O. cumana. 12

Within Orobanche spp., O. cumana is the only species that exhibits a clear race 13

structure with respect to sunflower genotypes. Single major genes (Or1 to Or5) were 14

reported to confer resistance to races A to E of O. cumana (Vrânceanu et al., 1986). Race 15

F, which overcomes the resistance gene Or5, was identified in Spain (Saavedra del Río et 16

al., 1994) and also in Romania (Pacureanu-Joita et al., 2008) in the mid 1990’s. Parasite 17

populations infecting sunflower genotypes that carry Or5 were later reported in Turkey 18

(Kaya et al., 2004). In Hungary, populations of O. cumana have been pathogenically 19

characterised as low to moderately virulent (races A to D) (Zoltán, 2001). 20

Different aggressiveness of populations of O. cumana race F have been reported 21

(Molinero-Ruiz et al., 2009) and genetic heterogeneity within some of these populations 22

has been suggested (Molinero-Ruiz et al., 2008). In fact, the L86 line, which was 23

registered as resistant to race F (Fernández-Martínez et al., 2004), showed up to 65% 24

incidence of very low infections, as compared with the susceptible control (3 and 23 O. 25

cumana stems per sunflower plant respectively) after inoculation with some race F 26

4

populations (Molinero-Ruiz et al., 2009). This suggested that these infections might be due 1

to highly virulent components (“new races”) within the parasite population. Therefore, 2

genetic diversity among and within populations of O. cumana race F might be responsible 3

for a) different aggressiveness on Or5, and/or b) consistent infections in sunflower inbred 4

lines with resistance to race F, respectively. 5

The random amplified polymorphic DNA-polymerase chain reaction (RAPD-PCR) 6

technique shows extensive divergence among, but little variation within, species. 7

Therefore, it is extensively used to distinguish biotypes and/or germplasm accessions 8

within weed species and crops (Volenberg et al., 2007; Zivkovic et al., 2012), as well as to 9

differentiate races or isolates of plant pathogens (Gómez-Lama Cabanás et al., 2012; 10

Villarino et al., 2012). Genetic diversity based on RAPD markers has been also assessed 11

among Orobanche spp. (Paran et al., 1997; Roman et al., 2003; Atanasova et al., 2005), 12

including O. cumana (Gagne et al., 1998). However, no molecular analyses have been 13

carried out to compare races or the infection ability of O. cumana populations on 14

sunflower genotypes carrying resistance genes. 15

In this work, the pathogenicity and molecular biology of highly virulent 16

populations of O. cumana from Hungary, Romania, Spain and Turkey were studied. The 17

objectives were: (i) to assess and compare the race or infection ability of O. cumana 18

populations on sunflower inbred lines carrying resistance to the parasite and (ii) to 19

investigate the genetic diversity among and within the most virulent populations of O. 20

cumana using RAPD analysis. 21

22

Materials and methods 23

24

Infectivity of highly virulent populations of O. cumana on resistant sunflower 25

Eleven populations of O. cumana were compared by their infectivity in three sunflower 26

inbred lines. High virulence was expected because they were collected from fields cropped 27

5

with sunflower allegedly carrying resistance to race E in Hungary, Romania, Spain and 1

Turkey (Table 1). In order to avoid previous cross pollination among them (Rodriguez-2

Ojeda et al., 2013), only those infested fields that were at least 5 km apart were considered 3

for the study. Orobanche cumana stems were collected at full maturity and seeds 4

recovered and kept until used, as previously described (Molinero-Ruiz et al., 2009). The 5

high virulence of the parasite populations was confirmed by inoculation on the sunflower 6

inbred line NR5 (Or5). NR5 is a tester line to differentiate races of O. cumana (differential) 7

that shows resistance to races A to E and susceptibility to populations of higher virulence 8

than E (Molinero-Ruiz et al., 2006). Populations of O. cumana were also inoculated onto 9

the differential sunflower inbred lines L86 and P96. Line L86 carries resistance to race F 10

and susceptibility to races E and G; P96 is resistant to races E and F but susceptible to race 11

G (Molinero-Ruiz et al., 2008) (Table 2). 12

Eight sunflower seedlings (replications) of each differential were inoculated with 13

each of the populations of O. cumana. Sunflower seeds were surface-sterilised by 14

immersing them in 10% sodium hypochlorite for 10 minutes, then thoroughly rinsed in 15

deionised water and incubated in the dark at saturation humidity in a germinator at 26±2ºC 16

until radicles were 2 to 5 mm long. Individual sunflower seedlings were transplanted into 17

pots with 175 g of soil mixture SS [sand:silt, 1:1, vol amended with 1 g l-1 of 16:7:15 (:2), 18

N:P:K (:Mg)] uniformly infested with 18 mg of parasite seeds and grown under conditions 19

conducive for infection (20 to 25ºC and photoperiod of 14 hours day-1). Fourteen-day-old 20

sunflower plants were transplanted, along with the infested soil, into 5 L pots containing 21

SS mixture and grown in the glasshouse at 10 to 32ºC for 10 additional weeks until 22

physiological maturity. Plants were watered as needed. 23

The degree of attack in sunflower genotypes (DA, i.e. the number of emerged O. 24

cumana stems per sunflower plant) was assessed weekly, from the emergence of the first 25

parasite stem until sunflower senescence. A standardised area under the DA progress curve 26

6

(SAUDAPC) was calculated by trapezoidal integration method, standardised by the 1

duration of parasite emergence in weeks (Campbell & Madden, 1990) and transformed 2

[sqrt (SAUDAPC+0.5)] prior to analysis of variance. When significant effects were 3

obtained, Fisher’s protected LSD tests (P = 0.05) were used for comparisons of inbred 4

lines, populations and their interaction. The experiment was performed twice and was set 5

up as a factorial on a completely randomised design. As no significant differences between 6

the two experiment replications were found for SAUDAPC (McIntosh, 1983), data were 7

pooled. 8

9

Molecular characterisation of highly virulent populations of O. cumana 10

To examine the molecular diversity among race F populations of O. cumana, stems of 27 11

populations from Hungary, Spain and Turkey, grown on sunflower NR5 in the 12

phenotypical characterisation described above and in previous studies (Molinero-Ruiz et 13

al., 2006; 2008; 2009), were collected. In order to assess the molecular diversity within 14

populations, stems of eight and four out of the 27 populations were also independently 15

collected from L86 and P96 respectively, and included in the molecular analyses. 16

Therefore, a final set of 39 samples of O. cumana was analysed using the RAPD-PCR 17

technique (Table 3). All stems were collected just before flowering and 2 to 3 cm long 18

apices were lyophilised until used. 19

Each sample consisted of total genomic DNA from three stem apices of each 20

population. DNA was purified using the Speedtools Plant DNA Extraction Kit (Biotools, 21

Madrid, Spain) according to the manufacturer’s instructions. Quality and concentration of 22

DNA samples were determined using a NanoDrop 1000 spectrophotometer (Thermo 23

Fisher Scientific Wilmington DE, USA). Finally, DNA samples were adjusted to a final 24

concentration of 25 ng/ μL and stored at –20°C until required for RAPD analysis. 25

7

Optimised PCR assays were carried out in a final volume of 25 μL containing: 0.6 1

μM primer, 200 μM dNTPs, 2.5 μl of 10× PCR buffer (800 mM tris- HCl, pH 8.3 to 8.4 at 2

25°C, 0.2% Tween 20 wt vol-1), 1.5 U of Ultratools Polymerase (Biotools, Madrid, Spain), 3

3.0 mM MgCl2, and 50 ng of parasite DNA. Amplification conditions were: 4 min 4

denaturation at 94°C; followed by 45 cycles of 1 min denaturation at 94°C, 1 min of 5

annealing at 37°C, and 2 min of extension at 72°C; and a final extension step of 6 min at 6

72°C. The temperature always varied at a rate of 5°C s-1. All reactions were done in a T1 7

Thermocycler (Whatman- Biometra, Goettingen, Germany). Amplification products were 8

separated by horizontal electrophoresis in 1.5% agarose gels containing 0.05 µl/ml 9

SafeView Nucleic Acid Stain (NBS Biologicals Ltd, Huntingdon, England) and visualised 10

over a UV light source. A 100- to 2,000-bp ladder (Dominion MBL, Córdoba, Spain) was 11

included in the electrophoresis. 12

RAPD analyses were carried out using 155 10-mer oligonucleotide primers 13

(Invitrogen Corporation, San Diego, CA, USA). Reactions yielding clear polymorphic 14

bands for all the samples were done at least twice and negative controls (no template 15

DNA) were included in each assay. Only primers producing consistent reproducible 16

polymorphism were considered. 17

A binary matrix based on presence (1) or absence (0) of amplicons of the same 18

molecular weight was generated for 18 selected primers. Genetic similarities among 19

isolates were calculated according to Jaccard’s similarity coefficient (Jaccard, 1908) and 20

dendrograms constructed from the distance matrices using the UPGMA method. The 21

robustness of each node of the dendrogram was calculated by generation of 1000 bootstrap 22

replications of the data (Nei & Kumar, 2000). Computations were performed using 23

Fingerprinting II Informatix Software (Bio-Rad Laboratories, Inc., Hercules, CA, USA). 24

The genetic structure of the populations was analysed by AMOVA (Analysis of Molecular 25

Variance) using Arlequin Software (Excoffier et al., 2005). Population structure proposed 26

8

was checked by Fixation index (FST) calculation and the corresponding significance P 1

value. In addition, pairwise FSTs were also calculated by pairs of populations using 2

genetic distance method as implemented in Arlequin Software. 3

4

Results 5

6

Infectivity of highly virulent populations of O. cumana on resistant sunflower 7

Both factors, population of O. cumana and sunflower inbred line, had a significant effect 8

(P < 0.001) on SAUDAPC. Population mean values of SAUDAPC ranged from 2.6 9

(average for LRB1603, OCR1, CU1102, CU200 and OCT3) to 13.6 (average for OCT4 10

and OCT2). When averaged across inbred lines, SAUDAPC values ranged from 2.4 in P96 11

to 10.3 in NR5 (Fig. 1). Remarkably, the SAUDAPC in each sunflower inbred line 12

depended on the population of O. cumana, as indicated by the significance (P < 0.001) of 13

the population x inbred line interaction. 14

There were populations identified as race E (OCR1, CU200 and CU1102), since 15

SAUDAPC was zero in NR5 and P96, and L86 showed 8.2 SAUDAPC (averaged across 16

populations) (Fig. 1). Race F was identified for populations OCH4, LRB1603 and OCT3, 17

since P96 showed a resistant reaction and high infection was observed in NR5 (12.3 18

SAUDAPC averaged across populations) (Fig. 1). Similar (population OCT3) or 19

significantly lower (populations OCH4 and LRB1603) infection occurred in L86 as 20

compared with NR5 (Fig. 1). When the inbred lines were inoculated with PA101 or OCT5, 21

a high infection occurred in NR5, L86 displayed a high (PA101) or moderate (OCT5) 22

reaction, and low infections levels were observed in P96 (Fig. 1). Finally, populations 23

OCT2, OCT4 and OCT6 from Turkey clearly overcame the resistance of the three inbred 24

lines and therefore they were identified as race G. The highest infection by these 25

populations occurred in NR5, and the infection in P96 was similar (OCT4 and OCT6) or 26

9

lower than the one in L86 (Fig. 1). In order to analyse the most virulent populations, stems 1

of OCH4, PA101, OCT2, OCT5 and OCT6 were collected from inbred line NR5. Also, 2

stems of populations PA101, OCT2, OCT5 and OCT6 were collected from lines L86 and 3

P96. All the collected stems were included in the subsequent molecular analysis, together 4

with stems collected in previous work by our research group (Table 3). 5

6

Molecular characterisation of highly virulent populations of O. cumana 7

Among the 155 primers initially assayed, 18 were selected since they produced 8

consistent polymorphisms. A total of 100 reproducible polymorphic bands were amplified 9

from the 39 assessed samples, the size ranging from 25 to 2,213 bp (N05 and G14 10

respectively). Amplicons generated by each primer varied between two (C16) and nine 11

(G03) (Table 4). 12

The dendrogram resulting from the UPGMA analysis of the RAPD data set 13

distinguished four well-differentiated clusters among the 39 samples of O. cumana (Fig. 14

2). Cluster I grouped the 24 samples collected in the south of Spain, which shared about 15

89% similarity. Two sub-clusters were identified within cluster I: one of them (Ia) grouped 16

exclusively N- samples of 16 populations, and the other (Ib) grouped N- and L- samples 17

from the remaining four populations from South Spain. The unique sample of O. cumana 18

from Hungary was grouped in cluster II, genetically closer to cluster I than to the rest of 19

the clusters. All parasite samples from Turkey shared 86% similarity and were grouped in 20

cluster III, irrespective of the sunflower inbred line they infected. Finally, the four samples 21

from populations collected in Central Spain grouped in cluster IV (80% similarity) (Fig. 22

2). 23

An AMOVA analysis among populations was carried out attending to geographic 24

origin of O. cumana (i.e. Spain vs Turkey vs Hungary). The analysis revealed that 60% of 25

genetic variability was due to differences among countries; however, up to 40% of the 26

10

genetic variability was due to differences among populations within the same country 1

(Table 5, analysis 1). A similar analysis but considering the South and the Centre of Spain 2

as different groups yielded the results showed in Table 5, analysis 2. The latter checked the 3

variability among the groups obtained in the UPGMA dendrogram (clusters I to IV, 4

including the isolate from Hungary as another cluster). In this analysis, percentage of 5

variation within groups was decreased down to 21%, presumably indicating that genetic 6

differences among Spanish groups (South vs Centre) accounted for 19% of the genetic 7

variability identified among populations of O. cumana in analysis 1. When pairwise 8

differences among populations were checked, there was almost as much divergence 9

between populations from South Spain and Turkey (78.3% of molecular variation) and 10

Centre Spain and Turkey (79.5%) as that between the Spanish geographical groups 11

(82.2%) (Table 6). 12

Concerning the molecular analysis within populations of O. cumana using 13

AMOVA, we found that genetic distance supported sub-clusters Ia and Ib from southern 14

Spain as genetically differentiated (Table 5, analysis 3). In contrast, differentiation 15

between samples of sub-cluster Ib obtained from lines NR5 (N-) and L86 (L-) was not 16

statistically supported (Table 5, analysis 4). In cluster III, we used AMOVA to analyse 17

whether samples obtained from P96 line (P-) could genetically be differentiated from N- 18

and L- samples, and whether N- and L- samples differed between them. The analyses 19

showed no significant differentiation (Table 5, analyses 5 and 6). 20

Finally, both presence and absence of bands were identified as markers for the molecular 21

groups in the UPGMA clustering, even for sub-groups Ia and Ib in cluster I (Table 22

7). Neither the presence nor the absence of any polymorphic band was associated with 23

genotypes NR5, L86 or P96, irrespective of the population of O. cumana (data not shown). 24

25

Discussion 26

11

Concerning the pathogenic characterisation of populations of O. cumana, some of them 1

were identified as race E, since they did not overcome Or5 gene in NR5 and were 2

pathogenic to L86 (Molinero-Ruiz et al., 2006; 2008). This was the case of OCR1 from 3

Romania and CU200 and CU1102 from Central Spain. Race F was identified in Hungary 4

(OCH4) for the first time, as well as in Turkey (OCT3 and OCT5), and confirmed in Spain 5

(PA101 and LRB1603). Populations of O. cumana causing the significantly highest 6

infections in P96 were all collected in Turkey (OCT2, OCT4 and OCT6). In accordance 7

with Kaya et al. (2004), who reported on the occurrence of one parasite race more virulent 8

than F in Turkey, our results allow the characterisation of OCT2, OCT4 and OCT6 as race 9

G. 10

The molecular analysis grouped O. cumana populations according to their 11

geographical origin. Moreover, parasite populations within close geographical areas could 12

be clearly differentiated as well (i.e. populations from Southern Spain vs. Central Spain). 13

This is in agreement with Pineda-Martos et al. (2013), who recently reported on two 14

clusters of O. cumana populations from Spain, one in Cuenca province (Central Spain) and 15

the other in the Guadalquivir Valley (Southern Spain). Cluster analyses using RAPD 16

markers also grouped, according to geographical regions, populations of crop species such 17

as Jatropha curcas (Rafii et al., 2012) and isolates of plant pathogens such as Verticillium 18

tricorpus (Usami et al., 2011) and Fusarium oxyporum f.sp. dianthi (Gómez-Lama 19

Cabanás et al., 2012). Gagne et al. (1998) previously studied eight populations of an 20

unknown race of O. cumana by RAPD analysis. They identified one cluster that grouped 21

populations from Bulgaria, Romania and Turkey, and another one corresponding to 22

populations from Spain. Our results confirm the molecular differentiation of populations 23

from Spain as compared with those from other countries in Europe. Moreover, populations 24

from Turkey were molecularly close to those from South Spain. 25

12

Populations of race F from South Spain were clustered irrespective of their 1

previously reported aggressiveness. Indeed, some populations identified as low, moderate 2

or highly virulent on NR5 (Molinero-Ruiz et al., 2009) were not genetically differentiated. 3

Only the highly aggressive CR202 and PG502 populations (Molinero-Ruiz et al., 2009), 4

grouped in the same sub-cluster (Ia). Qualitative resistance to O. cumana races A to E, 5

controlled by major genes and associated with complete absence of infection has been 6

reported in sunflower. In contrast, the genetic control of the resistance to race F into 7

sunflower germplasm has been described as both qualitative and quantitative (Fernández-8

Martínez et al., 2004; Pérez-Vich et al., 2006). In our molecular analysis within 9

populations, differences were found neither among N- and L- samples from CN503, 10

CO101, CT1503 and LR100 (sub-cluster Ia), nor among N-, L- and P- samples of OCT2, 11

OCT5 and OCT6 (cluster III) and PA101 (cluster IV), indicating that these populations 12

were genetically homogeneous. Therefore, the low infection levels found in L86 and P96 13

(Molinero-Ruiz et al., 2006; 2008) are not due to genetic heterogeneity within parasite 14

populations, but to the quantitative resistance of the inbred lines (Pérez-Vich et al., 2006; 15

Molinero-Ruiz et al., 2009). 16

On the other hand, race F populations could not be identified as a genetically 17

differentiated group, since they were present in all clusters after UPGMA analysis. These 18

results point to a polyphyletic origin for race F, arising independently at different 19

geographical origins, as has recently been suggested (Pineda-Martos et al., 2013). In 20

contrast, populations of race G were located only in cluster III, suggesting a monophyletic 21

origin for race G which should be further investigated. 22

Overall, and in spite of the disadvantages of the RAPD technique such as poor 23

reproducibility, lack of true homology between bands of the same length or dominant 24

inheritance (Mercado-Blanco et al., 2008), this molecular approach yielded very useful 25

results for our study, which was undertaken to search for polymorphisms among 26

13

taxonomically very close DNA samples [(i.e. polymorphisms within a race (F) of a weed 1

species (O. cumana)]. 2

This work constitutes the first geographical differentiation of O. cumana together 3

with pathogenicity and molecular traits inherent to each geographical group. Our results 4

indicate that molecularly homogeneous groups are clearly identified in highly virulent 5

populations of O. cumana, and that they are first related to geographical origin in Europe. 6

Then, some molecular groups relate only to one pathogenic group (i.e. race F from 7

Southern Spain) while others include different pathogenic groups from the same 8

geographical origin (i.e. races F and G from Turkey). Further research, based on more 9

robust molecular markers, might unravel the phylogeny of O. cumana. Orobanche cumana 10

and O. cernua are reported as one lineage of Orobanche species from the Old World 11

(Manen et al., 2004). In addition, our results provide useful information for the 12

development of molecular markers associated with geographical origin of O. cumana, 13

which could be used as a diagnostic tool for new parasite introductions into growing areas 14

where they might represent a threat to sunflower production. 15

16

Acknowledgements 17

Research supported by Ramón Areces Spanish Foundation, Spanish National Institute for 18

Agricultural Research (RTA04-048), Spanish National Research Council (CSIC) 19

(PIE200940I120) and Spanish Ministry for Science and Education (HH2005-0017). The 20

placement of S. Raranciuc was granted by the European Science Foundation (CA849). L. 21

Molinero-Ruiz was supported by an I3P post-doctoral contract (CSIC & European Social 22

Fund). We are grateful to J. Cobos, F. Gutiérrez and N. Lucena for technical assistance. 23

The critical reading of the manuscript and valuable suggestions prior to submission of J.M. 24

Fernández-Martínez and J. Mercado-Blanco are gratefully acknowledged. 25

26

14

References 1

2

ATANASOVA R, BATCHVAROVA R, TODOROVSKA E & ATANASSOV A (2005) 3

Molecular study of broomrape (Orobanche spp.) by RAPD analyses. Biotechnology & 4

Biotechnological Equipment 19, 51–60. 5

CAMPBELL CL & MADDEN LV (1990) Introduction to Plant Disease Epidemiology. 6

John Wiley & Sons, Inc., New York, USA. 7

DOMÍNGUEZ J (1996) Estimating effects on yield and other agronomic parameters in 8

sunflower hybrids infested with the new races of sunflower broomrape. In: Symposium 9

I: Disease tolerance in Sunflower (ed. A Pouzet), 118–123. The International 10

Sunflower Association, Beijing, China. 11

EXCOFFIER L, LAVAL G & SCHNEIDER S (2005) Arlequin ver. 3.0: An integrated 12

software package for population genetics data analysis. Evolutionary Bioinformatics 13

Online 1, 47–50. 14

FERNÁNDEZ-MARTÍNEZ J.M., PÉREZ-VICH B, AKHTOUCH B et al. (2004) 15

Registration of four sunflower germplasm lines resistant to race F of broomrape. Crop 16

Science 44, 1033–1034. 17

FOY CL, JAIN R & JACOBSON R (1989) Recent approaches for chemical control of 18

broomrape (Orobanche spp.). Reviews of Weed Science 4, 123–152. 19

GAGNE G, ROECKEL-DREVET P, GREZES-BESSET B et al. (1998) Study of the 20

variability and evolution of Orobanche cumana populations infesting sunflower in 21

different European countries. Theoretical and Applied Genetics 96, 1216–1222. 22

GÓMEZ-LAMA CABANÁS C, VALVERDE-CORREDOR A & PÉREZ-ARTÉS E 23

(2012) Molecular analysis of Spanish populations of Fusarium oxysporum f. sp. 24

dianthi demonstrates a high genetic diversity and identifies virulence groups in races 1 25

and 2 of the pathogen. European Journal of Plant Pathology 132, 561–576. 26

15

JACCARD P (1908). Nouvelles recherches sur la distribution florale. Bulletin de la 1

Société Vaudoise des Sciences Naturelles 44, 223–270. 2

JACOBSON R, GREENBERGER A, KATAN J, LEVI M & ALON H (1980) Control of 3

Egyptian broomrape (Orobanche aegyptiaca) and other weeds by means of solar 4

heating of the soil by polyethylene mulching. Weed Science 28, 312–316. 5

KAYA Y, EVCI G, PEKCAN V & GUCER T (2004) Determining new broomrape-6

infested areas, resistant lines and hybrids in Trakya region of Turkey. Helia 27, 211–7

218. 8

MANEN JF, HABASHI C, JEANMONOD D, PARK JM & SCHNEEWEISS GM (2004) 9

Phylogeny and intraspecific variability of holoparasitic Orobanche (Orobanchaceae) 10

inferred from plastid rbcL sequences. Molecular Phylogenetics and Evolution 33, 11

482–500. 12

MCINTOSH MS (1983) Analysis of combined experiments. Agronomy Journal 75, 153–13

155. 14

MERCADO-BLANCO J, COLLADO-ROMERO M & JIMÉNEZ DÍAZ RM (2008) 15

Estimación de la diversidad genética en hongos fitopatógenos mediante herramientas 16

moleculares. In: Herramientas Biotecnológicas en Fitopatología (eds. V Pallás, C 17

Escobar, P. Rodríguez-Palenzuela & JF Marcos), 167–188. Ediciones Mundi-Prensa, 18

Madrid, España. 19

MOLINERO-RUIZ ML, MELERO-VARA JM, GARCÍA-RUIZ R & DOMÍNGUEZ J 20

(2006) Pathogenic diversity within field populations of Orobanche cumana and 21

different reactions on sunflower genotypes. Weed Research 46, 462–469. 22

MOLINERO-RUIZ ML, PÉREZ-VICH B, PINEDA-MARTOS R & MELERO-VARA 23

JM (2008) Indigenous highly virulent accessions of the sunflower root parasitic weed 24

Orobanche cumana. Weed Research 48, 169–178. 25

16

MOLINERO-RUIZ ML, GARCÍA-RUIZ R, MELERO-VARA JM & DOMÍNGUEZ J 1

(2009) Orobanche cumana race F: performance of resistant sunflower hybrids and 2

aggressiveness of populations of the parasitic weed. Weed Research 49, 469–478. 3

NEI M & KUMAR S (2000) Accuracies and Statistical Tests of Phylogenetic Trees. In: 4

Molecular Evolution and Phylogenetics, (eds. M Nei & S Kumar), 165-186. Oxford 5

University Press, New York, USA. 6

PACUREANU-JOITA M, RARANCIUC S, STANCIU D, SAVA E & NASTASE D 7

(2008) Virulence and aggressiveness of sunflower broomrape (Orobanche cumana 8

Wallr.) populations, in Romania. Romanian Agricultural Research 25, 47–51. 9

PARAN I, GIDONI D & JACOBSOHN R (1997) Variation between and within 10

broomrape (Orobanche) species revealed by RAPD markers. Heredity 78, 68–74. 11

PÉREZ-VICH B, VELASCO L, MUÑOZ-RUZ J, DOMÍNGUEZ J & FERNÁNDEZ-12

MARTÍNEZ JM (2006) Registration of three sunflower germplasms with quantitative 13

resistance to race F of broomrape. Crop Science 46, 1406–1407. 14

PINEDA-MARTOS R, VELASCO L, FERNÁNDEZ-ESCOBAR J, FERNÁNDEZ-15

MARTÍNEZ JM & PÉREZ-VICH B (2013) Genetic diversity of Orobanche cumana 16

populations from Spain assessed using SSR markers. Weed Research 53, 279–289. 17

RAFII MY, SHABANIMOFRAD M, PUTERI EDAROYATI MW & LATIF MA (2012) 18

Analysis of the genetic diversity of physic nut, Jatropha curcas L. accessions using 19

RAPD markers. Molecular Biology Reporter 39, 6505–6511. 20

RODRÍGUEZ-OJEDA MI, FERNÁNDEZ-MARTÍNEZ JM, VELASCO L & PÉREZ-21

VICH B (2013) Extent of cross-fertilization in Orobanche cumana Wallr. Biologia 22

Plantarum 57, 559-562. 23

ROMAN B, ALFARO C, TORRES AM et al. (2003) Genetic relationship among 24

Orobanche species as revealed by RAPD analysis. Annals of Botany 91, 637–664. 25

17

SAAVEDRA DEL RÍO M, FERNÁNDEZ-MARTÍNEZ JM & MELERO-VARA JM 1

(1994) Virulence of populations of Orobanche cernua Loefl. attacking sunflower in 2

Spain. In: Biology and management of Orobanche. Proc. of the 3rd Int. Workshop on 3

Orobanche and related Striga research. (eds A.H. Pieterse, JAC Verkleijand and ST 4

Ter Borg), 139–141. The Royal Tropical Institute, Amsterdam, the Netherlands. 5

TAN S, EVANS RR, DAHMER ML, SINGH BK & SHANER DL (2005) Imidazolinone-6

tolerant crops: history, current status and future. Pest Management Science 61, 246–7

257. 8

USAMI T, KANTO T, INDERBITZIN P et al. (2011) Verticillium tricorpus causing 9

lettuce wilt in Japan differs genetically from California lettuce isolates. Journal of 10

General Plant Pathology 77, 17–23. 11

VILLARINO M, LARENA I, MARTINEZ F, MELGAREJO P & DE CAL A (2012) 12

Analysis of genetic diversity in Monilinia fructicola from de Ebro Valley in Spain 13

using ISSR and RAPD markers. European Journal of Plant Pathology 132, 511–524. 14

VOLENBERG DS, TRANEL PJ, HOLT JF et al. (2007) Assessment of two biotypes of 15

Solanum ptycanthum that differ in resistance levels to imazamox. Weed Research 47, 16

353–363. 17

VRÂNCEANU AV, TUDOR VA, STOENESCU FM & PIRVU N (1980) Virulence 18

groups of Orobanche cumana Wallr., differential hosts and resistance source genes in 19

sunflower. In: Proceedings of the 9th International Sunflower Conference (ed 20

International Sunflower Association), 74-82, LH/VPO, Torremolinos, Spain. 21

ZIVKOVIC B, RADOVIC J, SOKOLOVIC D, SILER B, BANJANAC T & 22

STRBANOVIC R (2012) Assessment of genetic diversity among alfalfa (Medicago 23

sativa L.) genotypes by morphometry, seed storage proteins and RAPD analysis. 24

Industrial Crops and Products 40, 285–291. 25

18

ZOLTÁN P (2001) Napraforgószádor - (Orobanche cumana Loefl.) populációk 1

patogenitásának vizsgálata. Növényvédelem 37, 173–182. 2

3

19

Table 1 Geographical origin (country, location and region) and year of collection of 11 1

populations of Orobanche cumana whose race or infectivity in genetically resistant 2

sunflower genotypes was characterised in this work 3

4

Location Population Country Location, region Year References *

OCH4 Hungary Bátonyterenye, Nógrád 2007 OCR1 Romania Braila, Braila 2005

CU200 Spain Segóbriga, Cuenca 2000 Molinero-Ruiz et al., 2008

CU1102 Spain Montalbo, Cuenca 2002 Molinero-Ruiz et al., 2008

PA101 Spain Palomares del Campo, Cuenca 2001

LRB1603 Spain La Rambla, Córdoba 2003 Molinero-Ruiz et al., 2008; 2009

OCT2 Turkey Cesmekolu, Kirklareli 2003

OCT3 Turkey Hayrabolu, Tekirdag 2003

OCT4 Turkey Malkara, Tekirdag 2003

OCT5 Turkey Muratli, Tekirdag 2003

OCT6 Turkey Babaeski, Kirklareli 2003

* References to previous works in which some populations were studied 5

6

20

Table 2 Reaction of differential sunflower inbred lines to races of Orobanche cumana * 1 2

Differential line Race of O. cumana

E F

G Spain C Spain S

NR5 R † S S S

L86 S S R ‡ S

P96 R R R S * Based on the results by Molinero-Ruiz et al. 2006 and 2008. 3

† R, resistant; S, susceptible. 4

‡ Inbred line L86, which was registered as R to race F, showed high incidences of very 5

few broomrape stems per sunflower plant when inoculated with populations of O. 6

cumana race F from the south of Spain, suggesting its quantitative resistance to the 7

parasite (Molinero-Ruiz et al., 2009). 8

9

21

Table 3 List of 39 samples of Orobanche cumana included in the molecular analysis of 1

populations of the parasite from Spain, Turkey and Hungary, and sunflower inbred lines 2

they were collected from 3

Sample Population of Orobanche

cumana, country *

Genotype of

sunflowerReferences †

N-OCH4 OCH4, Hungary NR5

N-MO197 MO197, Spain (C) NR5 Molinero-Ruiz et al., 2008

N-PA101 PA101, Spain (C) NR5

L-PA101 PA101, Spain (C) L86

P-PA101 PA101, Spain (C) P96

N-AA702 AA702, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-CN503 CN503, Spain (S) NR5 Molinero-Ruiz et al., 2009

L-CN503 CN503, Spain (S) L86 Molinero-Ruiz et al., 2009

N-CO101 CO101, Spain (S) NR5 Molinero-Ruiz et al., 2009

L-CO101 CO101, Spain (S) L86 Molinero-Ruiz et al., 2009

N-CO1002 CO1002, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-CP203 CP203, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-CR202 CR202, Spain (S) NR5 Molinero-Ruiz et al., 2006; 2009

N-CT803 CT803, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-CT1503 CT1503, Spain (S) NR5 Molinero-Ruiz et al., 2009

L-CT1503 CT1503, Spain (S) L86 Molinero-Ruiz et al., 2009

N-EC403 EC403, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-HE200 HE200, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-IN115 IN115, Spain (S) NR5 N-JC302 JC302, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-LC299 LC299, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-LG602 LG602, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-LR100 LR100, Spain (S) NR5 Molinero-Ruiz et al., 2006; 2009

L-LR100 LR100, Spain (S) L86 Molinero-Ruiz et al., 2006; 2009

N-MN802 MN802, Spain (S) NR5 N-PG502 PG502, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-SE500 SE500, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-SP102 SP102, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-TP402 TP402, Spain (S) NR5 Molinero-Ruiz et al., 2009

N-OCT2 OCT2, Turkey NR5 L- OCT2 OCT2, Turkey L86 P- OCT2 OCT2, Turkey P96 N-OCT3 OCT3, Turkey NR5 N-OCT5 OCT5, Turkey NR5 L-OCT5 OCT5, Turkey L86 P-OCT5 OCT5, Turkey P96 N- OCT6 OCT6, Turkey NR5 L- OCT6 OCT6, Turkey L86

22

P- OCT6 OCT6, Turkey P96 * The geographical origin Centre (C) or South (S) is indicated for samples from Spain. 1

† References to previous work in which some populations were studied. 2

3

23

Table 4 Consistent random amplified polymorphic DNA (RAPD) fragments generated for 1

39 samples of Orobanche cumana obtained from the sunflower inbred lines NR5, L86 or 2

P96 and screened with 18 DNA primers * 3

4

Primer code Sequence (5’-3’) No. of amplicons

scored †

Size of amplicons

(bp)

A17 GACCGCTTGT 5 526-1414

B08 GTCCACACGG 3 697-962 B19 ACCCCCGAAG 7 38-83 C11 AAAGCTGCGG 8 336-1112 C16 CACACTCCAG 2 305-644 G02 GGCACTGAGG 6 454-1057 G03 GAGCCCTCCA 9 331-1045 G04 AGCGTGTCTG 4 616-1189 G10 AGGGCCGTCT 4 522-1006 G13 CTCTCCGCCA 7 525-1416 G14 GGATGAGACC 8 400-2213 G15 ACTGGGACTC 6 495-850 N05 ACTGAACGCC 3 25-51 N09 TGCCGGCTTG 5 631-1161 N12 CACAGACACC 4 437-964 N16 AAGCGACCTG 7 473-1184 N17 CATTGGGGAG 5 528-1191 V15 CAGTGCCGGT 7 27-79 * Primers were tested twice on each of the 39 samples of O. cumana. 5

† Only intense and reproducible bands were scored. 6

7

24

Table 5 Genetic structure of 39 accessions of Orobanche cumana analysed using Analysis 1

of Molecular Variance (AMOVA) 2

Analysis * Source of

variation

d.f. † Sum of

squares

Variance

components

Percentage

of variation

Statistic

Probability

(P)

1 Among groups 2 151.927 8.614 Va 60.08 0.60084 P < 0.001

Within groups 36 206.021 5.722 Vb 39.92 - -

2 Among groups 3 250.499 11.365 Va 78.73 0.78732 P < 0.001

Within groups 35 107.450 3.070 Vb 21.27 - -

3 Among groups 1 25.125 2.179 Va 53.68 0.53676 P < 0.001

Within groups 22 41.375 1.881 Vb 46.32 - -

4 Among groups 1 0.75 0.000 Va 0.00 0.00000 P = 1

Within groups 6 4.5 0.750 Vb 100.00 - -

5 Among groups 1 5.581 0.626 Va 17.49 0.1749 P = 0.094

Within groups 8 23.619 2.952 Vb 82.51 - -

6 Among groups 1 4.702 0.462 Va 12.92 0.1292 P = 0.1594

Within groups 5 15.583 3.117 Vb 87.08 - -

* 1 = Spain vs Turkey vs Hungary, 2 = South Spain vs Centre Spain vs Turkey vs Hungary, 3 3

= South Spain Ia vs South Spain Ib, 4 = South Spain Ib N vs South Spain Ib L, 5 = Turkey 4

NL vs Turkey P, 6 = Turkey N vs Turkey L. 5

† d.f. = degrees of freedom. 6

7

25

Table 6 Comparisons of pairs of Orobanche cumana populations (FST 1

indexes) as a measure of population differentiation due to genetic structure 2

FST * P †

Comparisons according to geographic origin

Centre Spain vs South Spain 0.82240 <0.001

Centre Spain vs Turkey 0.79519 <0.001

South Spain vs Turkey 0.78309 <0.001

Comparisons according to infectivity on sunflower genotypes

South Spain N (Ia) vs South Spain NL (Ib) 0.53676 <0.001

South Spain N (Ib) vs South Spain L (Ib) 0.00000 0.991

Turkey NL vs Turkey P 0.17490 0.171

Turkey N vs Turkey L 0.12922 0.135

* Population pairwise FST index calculated using pairwise differences as 3

distance method as implemented in Arlequin software. 4

† Significance level of FST P values < 0.001. 5

6

7

26

Table 7 Markers of four molecular groups identified within 39 DNA samples of 1

Orobanche cumana after the molecular analysis of the band patterns obtained with 2

PCR amplification using 18 10-mer primers 3

Marker (bp)

Molecular groups

I II III IV

Ia Ib 27 - - - - + 32 + + + + -

37 - - - - +

41 - - + - +

48 + + + + -

51 + + + + -

57 - - - + +

62 - - + + +

73 + + - + -

79 - - + - -

331 - - - + -

336 - + - - -

400 - + - - -

458 - + - - -

463 - - - + -

525 - - - + -

526 + + + + -

616 - - - + +

631 + + + + -

697 + + + - +

715 - - - - +

754 - - - + -

823 + + - - +

953 - - - - +

962 + + + - +

1092 - - - - +

1112 + + + - -

1145 + + + - +

27

1161 - - - - +

1184 + - - - -

1416 - - - + -

1

2

28

Captions for Figures 1

2

Fig. 1 Reaction, expressed as the standardised area under the degree of attack progress 3

curve, of three sunflower inbred lines (NR5, L86 and P96), differentials for races of 4

Orobanche cumana, after inoculation with 11 populations of the parasite collected in 5

Hungary, Romania, Spain and Turkey and growth under shadehouse conditions (critical 6

least significant difference value = 5.15). 7

8

Fig. 2 Dendrogram derived from random amplified polymorphic DNA (RAPD) analysis 9

of 39 DNA samples of Orobanche cumana infecting sunflower (see Table 3) and 10

generated using 18 10-mer primers and the unweighted paired group method with 11

arithmetic averages (UPGMA). Scale represents percentage similarity using Jaccard’s 12

similarity coefficient. Bootstrap support percentages (≥75) for 1000 replicates are 13

indicated at the nodes. 14

15

29

Fig. 1 1

2

3

0

5

10

15

20

25

30NR5NR5

0

5

10

15

20

25

30L86

0

10

20

30P96

5

15

25

Sta

nd

ard

ised

area

un

der

the

deg

ree

of a

ttac

kp

rogr

ess

curv

e

Populations of O. cumana

30

Fig. 2 1

2

1

0.95

0.90

0.85

0.80

0.75

0.70

0.65

0.60

N‐CP203N‐EC403N‐CT803N‐CR202N‐CO1002N‐AA702N‐PG502N‐SE500N‐LG602N‐TP402N‐SP102N‐HE200N‐IN115N‐JC302N‐LC299N‐MN802L‐CN503N‐CN503N‐CO101N‐CT1503L‐CO101L‐CT1503L‐LR100N‐LR100N‐OCH4P‐OCT6P‐OCT5P‐OCT2L‐OCT2L‐OCT6N‐OCT2N‐OCT6L‐OCT5N‐OCT3N‐OCT5N‐PA101P‐PA101L‐PA101N‐MO197

80

100

100

83

100

77

II

III

IV

I

a

b